Changes in Health and Longevity over the 20th Century: Evidence from North Carolina
This project will use newly assembled vital-records microdata from 20th-century North Carolina to measure parent/child and sibling correlations in longevity. These records represent close to the universe of vital records (6.3 million births, 3.2 million deaths, and 1.35 million marriages) and cover the parents and siblings of many individuals aging in the later 20th century. Notably this pilot study includes men and women. This is possible because mothers' birth ("maiden") and married names are contained in their marriage and both birth records (in 86% of cases). Large sample sizes and the racial diversity of North
Carolina's population generate unprecedented linked samples for African Americans (in 1940 over 25% of the state's population). This differs from existing historical analyses which typically exclude women by necessity (e.g., the Union Army veterans) or because name changes at marriage preclude longitudinal surname linkage (e.g., census linkages). This study may be important for understanding the deteriorating relative health of U.S. women and contribute a deeper understanding of why U.S. women's longevity has failed to increase by as much as their brothers' in recent years. This study will help demonstrate the feasibility of using large-scale linked data, collect evidence on the quality of these data, and lay the basis for a larger grant proposal.
Funding Period: 07/01/2015 to 06/30/2016